I feel IA just adds to the pve Arcane problems. Between bugging priests for shields, druids for innervates and shamans for tides, becoming "that annoying mage" isn't really worth the measly raid dps boost.

Oh, no, I don't mean to become "that annoying mage." I personally hate asking people for things. I'm good friends with our shaman yet I almost never directly ask him for mana tide. So I haven't been demanding shields from disc priests. I just let them know I have the talent and stand near them when I can. They're going to shield you when you take damage anyway, most likely, whether you ask for it or not. I was kidding about needing to bribe them.

IA is still awesome in Ulduar with our own wards. I probably wouldn't have the points in it if I had to only rely on a disc priest.

I just tested with a priest, warlock, and mage. The first three tests are rolling IAs. The fourth test are separate IAs (letting the 10 seconds run out).

Shadow Priest, power word shield, shadow word pain

absorbed	spellpower	difference	total increase
-	2008	-	-
653	2105	+97	97
652	2174	+69	166
652	2222	+48	214
653	2255	+33	247
653	2279	+24	271
653	2296	+17	288

Warlock, curse of agony, damage spread out 4 ticks at a time which is why the fifth tick spiked slightly.

absorbed	spellpower	difference	total increase
-	2008	-	-
465	2077	+69	69
466	2133	+56	125
466	2178	+45	170
466	2213	+35	205
563	2257	+44	249
524	2286	+29	278

Mage, scorch <10 sec apart

absorbed	spellpower	difference	total increase
-	2008	-	-
2202	2338	+330	330
1283	2477	+139	469
1219	2583	+106	575
1278	2707	+124	699

Warlock, shadowbolt >10 sec apart

absorbed	spellpower	difference	total increase
-	2008	-	-
2950	2450	+442	442
-	2008	-	-
2472	2133	+464	464
-	2008	-	-
2950	2450	+442	442

Observations:

• The first "tick" in a rolling IA always gets the correct, full 15% bonus spellpower. However, it rounds down.
• The original theory that each rolling IA is on its own separate but stacking 10sec timer is incorrect.

---


Edit:

We now know how each additional rolling IA is calculated to form your current spellpower value:

[New SP Bonus] = [Absorbed Damage] * 0.15 + [Old SP Bonus] * ([Remaining Buff Time] / 10)

Taking the first table, for example:

Shadow word pain, ticks are 3 seconds apart

Absorbed	Spellpower	Difference	New SP Bonus	Calculation
-	2008	-	-	-
653	2105	+97	97	97 = 653(.15)
652	2174	+69	166	166 = 652(.15) + 97(7/10)
652	2222	+48	214	214 = 652(.15) + 166(.7)
653	2255	+33	247	247 = 653(.15) + 214(.7)
653	2279	+24	271	271 = 653(.15) + 247(.7)
653	2296	+17	288	288 = 653(.15) + 271(.7)

Here I'm addressing the question of whether or not one should build up stacks of Arcane Blast before using his or her 4pcT8 proc (a second free MBarr after the first MBarr). I've seen more than one person state recently that an Arcane Mage should first build up three stacks of AB before using the 4pcT8 proc, and I do not see this being the case at all.

I preface by saying I'm not a statistician, so I may have botched some of this. Feel free to correct me if so:

Assume a standard Arcane build with the Arcane Blast Glyph.

Missile Barrage proc rate = 20% per Arcane Blast (or 80% chance not to see it per AB).

Three Arcane Blasts in a row means (0.8*0.8*0.8 = 0.512) 51.2% chance of not seeing it proc, or a 48.8% chance of seeing it proc. (This means you'll lose 48.8% of your 4pcT8 procs to a normal MBarr proc when you're building your three Arcane Blast stacks.)

3xAB including the Glyph of Arcane Blast increases the spell damage on a MBarr proc by 54%.

Since all other variables are equal, we can just set your typical AM damage as the variable D.

I think it's then sound to multiply the normal MBarr damage by 0.512 to account for the fact that only 51.2% of the proc MBarrs will make it through three Arcane Blasts without a reproc of MBarr.

So, by casting ABx3, you're changing the base value of your 4pcT8 proc to 0.512D.

You're then gaining 54% additional damage on top of that 0.512D, making it 0.788D.

Now I want to ask whether casting ABx1 or ABx2 benefits us by doing the same math:

ABx1 = 80% chance of not seeing a new MBarr proc.
Lose 20% of 4pcT8 procs to a new MBarr proc.
4pcT8 MBarr proc will be worth 0.8D with a ABx1 buff giving it 18% additional damage, making it finally worth .944D

ABx2 = 64% chance of not seeing a new MBarr proc.
4pcT8 MBarr base = 0.64D with 36% bonus damage, making it worth .87D.

So my conclusion is that when an Arcane Mage has a 4pcT8 proc off MBarr, one gets the most damage yield by immediately using the MBarr proc as opposed putting up any AB stacks first.

Patchinko, try using the same logic to prove that it's better to use MBarr before 3 stack even without 4T8, it'll probably work.

What you're asking is whether using MBarr as soon as it procs as opposed to waiting until after a 3xAB stack is worth more damage on the MBarr side.

Assume MBarr procs off your first AB. Your MBarr damage will be 1.18D if you use it that moment.

Cast your second AB. There's an 80% chance you won't proc MBarr. Your MBarr damage after the second will be (0.8D)*1.36 = 1.09D.

If you cast a second and a third AB, there's a 64% chance you won't proc MBarr. Your MBarr damage will be (0.64D)*1.54 = .99D.

Now assume MBarr procs off your second AB as opposed to your first. MBarr damage will be 1.36D if you use it that moment.

If you cast a third AB, damage becomes (0.8D)*1.54 = 1.23D.

Basically, what this says is that if you use MBarr as soon as it procs, you'll get the most damage out of your MBarr. This is because if you proc MBarr again while having the MBarr proc already on you, you lose the potential damage from the original MBarr since you didn't cast it at all.

Like you point out below, in a cycle, the added damage on your ABs makes a 3xAB/MBarr rotation higher damage at the end of the day.

That is not correct. Using MBarr at 3 stack is higher dps. I would suggest checking the following:

The Arcane thread
Mathematics of dynamic cycles

You're right, I shouldn't have made that statement. Your point that one should consider the entire cycle is key there--the increased damage on the extra ABs will make up the difference in the MBarr damage. I'll fix my post to reflect that.

That's correct, I tested it with some spreadsheeting and confirmed it with his (as best as I could) and my numbers.
The formula would be something like:
[New SP Bonus] = [Absorbed Damage] * 0.15 + [Old SP Bonus] / 10 * [Remaining Buff Time]

So for the sake of simplicity: The spellpower buff is still only present over 10 seconds, or at least normalized to 10 seconds.

I tried a IA 3/3 build on Hodir last night

with talented glyphed frost ward
and talented glyphed mana shield
plus 2/2 magic absorbtion

survivability was really good, resisting / absorbing most of frost damage.
Not sure about dps tho.

Build was pretty fine in pvp

As i noticed from logs and heard from other persons, the glyphs for frost/fire wards don't work with area damage.
From my point of view the glyphs are useless in Ulduar, because the damage in most encounters is area damage, perhaps only the fireballs from razorscale will be reflected.

I think you should read the tooltips again... Fire Ward & Frost Ward. Both are currently a chance to reflect, not absorb more... and unfortunately they don't work against AOE, as noted previously by others.

The idea behind an IA build is that you might not have a disc priest casting PW:S on you all the time. For fights with fire or frost damage, IA Arcane mages spec out of Ice Shards and pick up Frost Warding:

Frost Warding - Spell - World of Warcraft

which gives your wards a 30% absorb. The question above was whether or not the glyphs increased this absorption percentage.

Yes the point i figured out was only that the both glyphs are nearly with no effect in ulduar. IA is a nice and good talent in ulduar.

I have a combat log of a duel between myself and two L80 Shaman. I use a glyphed talented Fire Ward (Frost Warding 2/2) and have no talents in Molten Shields.

Searing Totem Rank 1

111 Absorbed
40 Frost Warding Procs (44.44%)
3 Reflect

Magma Totem Rank 1

114 Absorbed
34 Frost Warding Procs (29.83%)

So the AoE checks out as predicted. Single target seems high for the sample, will try and add more later.

If anyone else has any data PM me.

Hi, maybe silly question, but just to confirm haste>crit for arcane yes?
Same with FB/TTW I see.

Edit: woops i just discovered the search option never mind me >.<

Quick question, is there anyone working on an up-to-date frist page ? The one we have dates pre-3.1 or is a mix of evrything.... BIS items and stuff like that ? Just curious

Just use Rawr to determine what gear is best for your current set up. Also a few pages back there's some discussion as to what is BiS for Arcane.

When is the best time to use arcane power? In rawr heroism is always used at the best moment but my guild never really has predictable heroisms. So, should I try to stack cooldowns or just use everything as soon as I can?

It'd probably be best to simply toggle off heroism then for the purposes of finding how to sync cooldowns. When I turn it off it suggests to wait for either mana gem, mana gem + trinkets, or trinkets, depending on gear setup. I.e., only use AP when you can use it with something else.

Common agreement is that heroism should be used as early as possible where practical (boss is in position, tank has solid threat, all debuffs/buffs are up etc) to get the boss to execute range faster.