I was under the impression the rotation is ABx4 MB, repeat "unless MB doesn't proc"
When that's the case then you just keep spamming AB until you get the proc and then MB, right?

It looks like you two are talking about the same thing. The terminology Rawr and many here use is ABSpam4MBAM. That means spam AB until you have both a 4 stack and MB proc and then use MBAM.

Hi guys !

You may wonder why I just post here as i'm actually not playing mage as main character, but i spent almost all of my afternoon discussing the value of the 2 pieces Tier 10 bonus with my guild's best mage and I think our work could be of use here. I apologize in advance for my english, I just hope it's good enough for you to understand me !

So here we go analyzing the value of 2T10 : 12% spell haste for 5 second after channeling Arcane Missile.



Calculus principle :

We need to calculate the percentage of spells that are gonna benefit of this bonus.

In order to do so, we'll calculate how many spells are affected each "sequence" (4+ * Arcan Blast (AB) then Arcane Missile (AM)) [this number always stay the same and only depends on your haste], and the number of spells cast per sequence [this one is random and depends on your luck at getting the AM proc, we'll average that].



Number of spells cast under 2T10 per sequence :

Only haste at the beginning of a cast matters as the "speed" of a cast won't change while casting, it will use the haste value the caster had at the beginning of the cast.

With only few haste, AB cast time is little under 2.5 sec (let's say 2.4 sec), which allows you to cast 3 AB under 2T10 (2.4 + 2.4 = 4.8 sec and 0.2 sec remains for you to start casting the third AB).

Increasing haste from here won't increase the effect of the T10 bonus as it will keep being active on only 3 AB, it'ill just last longer while 3rd AB is casted, but that one's fully affected anyway (as explained earlier).

However, when reaching the 50% haste cap (raid buffed), you get to cast 4 AB under 2T10 effect [3*2.5/(1+50%) = 5 seconds] : 3 AB fully under the buff effect and some remaining milliseconds to start casting the 4th. Of course, you'll have to make sure you get to cast this 4th in time, i.e. you need some margin to make up for latency.

Second haste cap is 100% [4*2.5/(1+100%) = 5 seconds], which is unreachable except under cooldowns effects (like Blood Lust or Icy Veins).

How much haste you need to achieve the 50% haste cap ?

You get 5% from the Shaman totem, 3% from the Ret/Moonkin aura, 12% from 2T10 and 6% from talents. So (1+haste/100)*1.05*1.03*1.12*1.06 = (1+50%) resulting in 16,82% haste (character sheet unbuffed) or 552 haste rating (plus the margin discussed earlier that depends on your latency).

Under the effect of Blood Lust, you need 19.82% haste (unbuffed character sheet) - 650 haste rating - to reach the 100% cap.

You can also reach the 100% without Blood Lust using Icy Veins, but then you'll need 29.80 haste - 978 haste rating.

Without those cooldowns, the 100% cap can be reached with 55,76% haste (1830 haste rating), which for now can be reached using a haste potion (+500) and the trinket "Scale of fate" (+432) at the same time (thus needing only 898 haste rating - 27.39%).

Don't ever exceed 150% haste (which will happen under Blood Lust + Icy Veins if you have 814+ haste rating - 21.81%+ haste unbuff character sheet), you're then under gcd.



Average number of spells cast each sequence :

Calculus assumption : As you need some time to react to the AM proc, if that proc happens on the 4th (or aly later) AB, you'll still cast one more AB (5th one in thit case) before AM.

We'll calculate the probability of the different possible situations :

- 5 spells sequence (optimal : 4 AB + 1 AM, proc must happen on one of the 3 first AB) : 1-0.6^3 = 78,4 %
- N spells sequence (N > 5 : N-1 AB + 1 AM, proc must happen on the N-2th AB) : 0.6^(N-3)*0.4

So the average number of spells cast each sequence is 5*(1-0.6^3) + 6*(0.4*0.6^3) + 7*(0.4*0.6^4) + 8*(0.4* 0.6^5) + ...

Calculus done, you cast 5.54 spells per sequence in average.



Conclusion : uptime and efficiency

Depending on you raid haste, the value of this 2T10 bonus can be (equivalent haste rating depends on your actual haste, i'll give it for necroma's 27,3% haste unbuffed for yoou to make yourself an idea) :

- Between 0% and 50% haste -----> 54.15% uptime / 6.50% average haste (equivalent to 8,27% character sheet unbuffed or 271 rating)
- Between 50% and 100% haste --> 72.20% uptime / 8.66% average haste (equivalent to 11,02% character sheet unbuffed or 361 rating)
- Over 100% haste -----------------> 90.25% uptime / 11.07 average haste (equivalent to 14,09% character sheet unbuffed or 462 rating)

4T9 bonus is 5% crit (226 rating), and assuming haste is between 1.5 and 2 times better than crit (for necroma - our guild's reference mage, and according to rawr - exactly 1.62 times better), so 2T10 is about 2.59 times more interesting than 4T9 for a HL mage (that should reach the 50% cap).

Enjoy Tier 10 !



Arcturus (Paladin) & Necroma (Mage) @ Armada - Elune - EU

That's some nice TC Pongueur. What I'm taking from this is: In addition to being aware of the 50/100% steps, stack your CDs wisely to avoid hitting the t10-procced GCD cap.

Haste multipliers from talents/buffs/t10 proc/Black Magic enchant: 1.05*1.03*1.12*1.06*1.07624 = 1.3818%. So you'll hit the 150% cap with (1+Haste)*1.3818=1+150%. We then have Haste Cap = (2.5/Multiplier) - 1. Haste would then have to be 80.92%, or the equivalent of (80.92*32.79) = 2653 rating.

BL = 30%
IV = 20%
Haste pot = 400 rating = 12.199%
Scale of Fates = 432 rating = 13.175%
Black Magic = 250 rating = 7.624%

Soooo, keeping in mind that all haste buffs are multiplicative, any haste over 15.98%, or 524 rating, will be wasted under BL+IV+BM as it's faster than the GCD cap of 1 second. So basically any other two CDs are fine when used at the same time until we get deep into heroic ICC gear.

Assuming full raid buffs and the Black Magic enchant:

Haste rating cap	Without t10	2pc t10
Base	3365	2653
BL	1832	1284
BL + IV	980	524
BL + pot	1276	788
BL + IV + pot	490	87
IV	2258	1665
IV + pot	1656	1127
Pot	2643	2008

Nice theorycrafting, but I believe that you are overstating the impact of "wasting" haste due to the soft cap:

Even when I'm geared at 900 haste, I seriously doubt that staggering your cooldowns is always worthwhile. Assuming you get an opportunity to blow heroism at the start of the fight, one could spend up to 15 seconds fishing for a ring proc, before blowing IV/AP/wild magic potion. Sure there is an extent of waste, by going over haste cap and using a suboptimal potion, but that should be more than offset by the actual benefits of buffing your haste up to the very limit while using DPS cooldowns.

I'd be interested in more calculations on not only when I start hitting the cap, but when it actually stops being worthwhile to pop IV regardless, to get more bang for my buck with my other cooldowns. Hell, even without popping a single DPS cooldown, the soft haste cap is exaggerated. Consider that we're only touching the cap because of multiplicative haste stacking, which in effect means that IV gives a stronger DPS bonus during those conditions than any other time.

I think we can arrive at numbers that are way more useful in reality, than knowing when you start touching the cap. Also, does anyone have any decent numbers on haste pot vs. wild magic potion? Particularly when stacking cooldowns.

Then there's also the black magic enchant, and the hyperspeed accelerators for engineers. The accelerators are a no-brainer, but personally I'm really not sold on the black magic enchant, because it potentially wastes a cooldown during heroism, when +63 SP would scale extremely well if I'm stacking cooldowns. I'm guessing it would turn into a pretty close call overall though, and I'm very interested if anyone has numbers on it.

We've spent the last couple of days working on Professor Putricide on both 10 and 25 mans, and I notice Im having a hell of a time on the slimes, specifically the Volatile Ooze. Our raid leaders do not want us moving into range to dps the ooze prior to him selecting a target. Once that happens, we're to dps until the last second, and then move onto the targetted person. This seems to give me roughly 2 casts, 3 if pom is up before I have to move.

With the lower range of arcane, and the having to wait, my V. Ooze damage is unacceptably low, normally running on the low end of everyone else in the encounter. I can offset this a bit by saving my cooldowns for those ooze specifically, but obviously not on every one.

Anyone have any tips for how you are handling them that can improve my performance on the ooze?

Edit - I should clarify where we are standing a bit. We move into the far opposite corner when the slime appears, prior to him targetting anyone. This being the case, that means if I am not the target, I have to move to damn near the center of the room to be within range. By the time i get within range to cast, I get one AB off, and have to move back to stack onto the targetted person. I might get ABx3 total if im lucky while moving twice, but never enough dps time for a full stack and an AM.

Calculations made for a 900 haste rating mage, i'm trying to compare using IV in or out of BL.



1/ Blood with IV and arcane power then Haste Pot, then Trinket (scales of fate) under black magic proc. (Arcane Power dps gain evaluated to 15% stead of 20% due to bad stacking with AB stackable buff according to a coming post).

You get actually 11,07% haste from 2T10 under BL (over 100% cap) and 8,66% out, and you will go over gcd haste cap uner BL + IV (150%). You'll be under 100% cap out of BL so 2T10 is worth 8,66% haste.

BL : Haste multiplication : 1,3*1,05*1,03*1,06*1,1107 = 1,6552
No BL : Haste multiplication : 1,05*1,03*1,06*1,0866 = 1,2457

Averaged Black magic proc that will happen (once exactly) under BL is worth 63 (=250*10/40) haste rating when not hard capped.

15 sec : BL + IV + Arcane Power : 2,5*1,15*(1,1107/1,12) = 2,8511
5 sec : BL + IV : 2,5*(1,1107/1,12) = 2,4792
15 sec : BL + Pot : (1+(963+500)/3279)*1,6552 = 2,3937
5 sec : BL : (1+(963)/3279)*1,6552 = 2,1413
10 sec : Trinket + Black Magic : (1+(900+432+250)/3279)*1,2457 = 1,8467
10 sac : Trinket : (1+(900+432)/3279)*1,2457 = 1,7517

60 sec : Average : ((2,8511*15)+(2,4792*5)+(2,3937*15)+(2,1413*5)+(1,8467*10)+(1,7517*10))/60 = 2,2960

Average DPS during this minute is 129,6% higher that normal cycle DPS.



2/ Blood with Haste Pot then Trinket, then IV on Black magic proc.

You will just hit the gcd cap if black magic procs under Blood+Pot but we can consider no haste is lost and average black magic proc (63 rating). You'll be over the "100% cap" all time except under IV without Black magic so 2T10 is worth 11,07% haste.

BL : Haste multiplication : 1,3*1,05*1,03*1,06*1,1107 = 1,6552
No BL : Haste multiplication : 1,05*1,03*1,06*1,1107*1,2 = 1,5280

15 sec : BL + Pot + Arcane Power : (1+(963+500)/3279)*1,6552*1,15 = 2,7528
5 sec : BL + Pot : (1+(963+500)/3279)*1,6552 = 2,3937
20 sec : BL + Trinket : (1+(963+432)/3279)*1,6552 = 2,3594
10 sec : IV + Black magic : (1+(900+250)/3279)*1,5280 = 2,0639
10 sec : IV : (1+(900)/3279)*1,05*1,03*1,06*1,0866*1,2 = 1,9051

60 sec : Average : ((2,7528*15)+(2,3937*5)+(2,3594*20)+(2,0639*10)+(1,9051*10))/60 = 2,3356

Average DPS during this minute is 133,56% higher that normal cycle DPS.


According to these calculations, it seems better not tu use IV under BL, but out of it on a black magic proc, but the difference is so slight that using cooldowns quicker can be of more use...

Arcane Power doesn't increase your damage by 20%. It stacks additively with AB stacks (and some other talents) and thus is closer to a 15% damage increase.

At first some comments on the t10 Bonus. From just reading the setbonuses some mages may think the 2set bonus is weak, but from testing it myself and pongeur's math proved my results, it's a pretty strong bonus. With an average amount of haste all 4 Arcane Blast will benefit of the 12%.

Concerning Bloodlust. Pongueur your math should be correct. However, you cannot "time" your passive procs. The only time where you can really expect your procs is at the beginning of a fight.
And even, if you see for example, black magic has expired and now need to use haste pot to maintain 1,00 Arcane Blast casts you can still have an embrace of the spider proc (or any other passive haste proc) which destroys your optimal Rota.

I want to substantiate my way of thinking about this topic with a comparison:

Embrace of the spider vs Abyssal Rune (210,59 Dps both on maxdps calculations so for the moment, dpswise they are almost equal for arcane)

If you are at 1,00 AB with Icy Veins+Heroism and get a proc of Embrace you will be 505 Haste below the Hastecap. However if you run with Abyssal Rune and modify your gear (84 passive haste), you will have the constant +haste and no proc problem.
This example should show that when getting close to the hastecap, hasteprocs/pots will be a dps loss compared to the equal trinkets with spellprocs/pots.
With 800 haste, managing some of the procs could be viable (i doubt you can do it in a bosssituation). But with 1100+ haste the situation is more complex.


There are 2 suggestions to counteract this.

1) Reduce your haste by replacing equal non-haste gear so that your procs will bring you to 1,00 and not above and manage your procs/trinkets to substain 1,00 casts.
2) Replace your hasteprocs/pots with equal dps effects.

to 1) you cannot replace hasteitems. SPI+SP items are lower dps, SP+Crit items are lower dps.
2) Is the way to go, if it's possible by the gear/items/procs available. (something like Reign is pretty optimal cause it will never cap)

The increase of haste cannot be stopped with the current gear progression and haste will outweight crit more, the longer the fight (including bloodlust) will be.

I don't think a proc haste trinket's not a good idea, especially cause you can't control when to activate it... and so you waste it sometimes, or you get it when you just don"t want of it.

I noticed that T102p proc was allowing for .8/.9 sec AB (according to quartz)...I think it allows you to shorten the GCD.

Buffs:
Hyper Accelerators
BL
Black Magic
Scale of Fates
IV

Just providing this info incase nobody else has noticed this.

(still using them all, pulling IV out of the sequence for BL)