Robes of the Elder Scribes have an internal cooldown of 45-50 seconds. This information has been posted before on these forums, indeed even in this thread.

Were the cooldown to be removed, you are correct that the Robes would be a very strong fit in any situation where AM is used, similar to the Wrath of Cenarius and Hyjal exalted rings. Indeed the hyjal exalted ring's proc is worth over 60 passive damage when using MSD/Ashtongue and AM spam.

I had a conversation with a hunter a few weeks ago, who claimed that threat is reduced by range in 3 steps: mele range, mid range, and long range. Previously I had been of the impression that threat was full at melee range, and 20% less outside of melee range. Do any of you know how threat is affected by range? The reason I ask is this- I know with fire you can go to 42yds, which is a lot further than 30yds of arcane, and if 30yds is 'mid range' would it not have less threat reduction than 42?

That's pretty much completely wrong. Here's how range affects threat:

The tank (current aggro target) has what we will call 100% threat. In order for the mob to retarget onto a player inside melee range, that player needs to gain 110% of the tank's current threat. For a mob to retarget onto a ranged player, that player needs to reach 130% of the tank's threat. So ranged classes have a larger buffer before they pull aggro than melee classes, but the actual threat generation is unaffected.

Thanks for the clairifcation on pulling aggro and actual threat values. Are you certain that all ranges beyond melee pull at 130% threat?

While that is correct, it's still wise to not pass tank's threat even as a ranged class. Because if another ranged class pulls aggro, the tank is no longer the 'tank' and that boss/mob is just going to run wild gibbing people until the tank is back on top of the threat list. But if you DO happen to pass the tank in threat, you have a larger buffer than a melee class where you can either invis or stop threat generation and wait for the tank to pass you again.

Just to beat the obviously dead equine - the 130%/110% threshold is dependent only on your distance, NOT on the type of damage you deal. You can be a mage, stand next to a mob, throw nothing but ice lance, and pull at 110%. In fact, I have heard (but not personally verified) that you could be at max range, build 110-129% threat, then do nothing else but run toward it, and pull aggro as soon as you come into melee range.

It would be nice if I could simply refer people to wowwiki.com for things like this, and I shall, since much of the math on it is valid. But as with much of the Internet, you can't believe everything you read on it, either, so caveat lector.

This is also big argument in the Mages vs Warlocks threads when it is argued that mages, the "kings of AoE", aren't as good as Warlocks because SoC has a 30% threat buffer and Arcane Explosion/Blastwave/CoC/Dragon's Breath are 10%.

So Warlocks can put out more damage before pulling.

Blink to the locks when you pull.

I did some theory crafting yesterday afternoon and posted my results on my guild's forums. I wanted to get some more professional feedback on it since people here seem to be knocking Arcane Missiles spam.
***
Attached is a spreadsheet to show how mages get their mana back and how important certain things are to us, JoW and Shadow Priests for example.

It should be noted that in a mana burn (AB spam, AE spam), mana tide and innervate is more important because our mana pool is the limiting factor and that usually means there is a goal to be done in a certain amount of time (AoEing Kael's weapons, bursting RoS). But over the course of a fight they really don't help us all that much.

The biggest contributors are Shadow Priest, Spirit/Mage Armor, and Judgement of Wisdom.

Note: JoW up 100% of the time in a 6 minute fight would cost the paladin 4860 mana, or just slap the mob every 20 seconds.

With BT gear, and assuming fully raid buffed/potted, an Arcane Mage spamming Arcane Missiles for 6 minutes straight does 1476dps

Also assuming 433 raid buffed Spirit.

The math I used is as follows:
Arcane Missiles = 740mana + 6% = 784 mana
-56,476 mana (72 volleys in 6 minutes - No MSD procs)
+5,647 mana w/ Arcane Concentration 7.2 volleys @ no mana
+12,000 mana pool
+4,000 mana from pots (2)
+2,300 mana from gems (emerald/ruby)
+7,272 mana from evocate (32.5+0.625x433) / 5 x 2 x 4 x 15 --- 9,090 with 2pcT6
+3,542 mana from Blessing of Wisdom (49*360seconds/5)
+9,821 mana from spirit/mp5 0.45x(32.5 + 0.625x433) = 136.4mp5 x 360 / 5
+18,000 mana from Shadowpriest with 1000 dps (1000*0.05*360)
--------------------------------------------------------------------------------
6106 mana left

+13,320 mana from 100% JoW (72 Volleys x 5 Missiles x 50% chance x 74 mana)
+2,160 mana from untalented mana spring (360/2 x 12)
+2,880 mana from Mana Tide (24% of 12,000 mana)
+4,850 mana from Innervate (32.5+433x0.625)x4x4



With full buffs, pots, gems, evocation, and 1000dps shadow priest alone, AM spam is sustainable and you would have +6106 mana left after 6 minutes, with your potion and gem timer up. This left over mana, and mana from all other abilities mentioned would allow for a AB spam burn, or help account for MSD.

Things NOT taken into consideration in the above:
Resists
Partial Resists
Bloodlust (increases mana usage by 2688 per Bloodlust, AM becomes 3.5s cast = 11.4 volleys as opposed to 8)
Focus proc from Mystical Skyfire Diamond
Arcane Power

Things assumed 100% active (yes I know 4pcT6 makes this even more favorable):
Misery
CoS
Wrath of Air Totem
4pcT5
Exalted Hyjal Ring
Crusade Deck
The Lightning Capacitor

How much damage is all of this?
In the "best" gear available... fully buffed
Spell Damage this is around 1800 spell damage
AM coefficient = 187%
Effective 3366 spell damage
Per Missile 3366 / 5 = 673.2 dmg from Spell Damage + 260 base = 933.2 per missile.
Now add in CoS and Misery and Arcane Instability: 933.2 x 1.1 x 1.05 x 1.03 = 1110 per missile

Over a 6 minute fight, 72 volleys x 5 missiles = 360 Missiles

I'm going to assume a 33% crit rate, Very conservative for gear like this, it would be higher.
120 Missiles Crit
240 Missiles Non Crit

Crit Dmg = 1110 x 1.75 = 1942.82

TLC Dmg = 120 / 3 = 40 x 800 = 32000

TOTAL Damage = (120x1942.82) + (240x1110) + 32000 = 531538 Damage
(TLC is 6% of this damage for those curious)
Also 7382 damage per volley on average

531538/360 = 1476.5 DPS

3.6 of those missiles will be resists.
A lot will be partial resists and I don't know how to model this but I would think MSD/Bloodlust would offset this.

As far as how to model MSD... I know this isn't accurate, but its close I think..
6 chances per volley = 432 chances at 5% each = 21 procs

Or you can model it on chance per volley..
1-(0.95^6) = 26.5% chance per volley x 72 volleys = 19 procs.. so its close either eay.

Let's call it 20 procs in the fight. 20 of those Volleys will take half the time, leaving 2.5 x 20 = 50 seconds of extra time to cast 10 more volleys, of those 2 will proc focus, giving you time for one final volley :P

72 + 10 + 1 = 83 total volleys with MSD
11 x 7382 = 81202 extra damage from MSD, at the cost of 11 x 784 x 0.9 = 7761 mana

Or, MSD adds 81202 / 360 = 225.5 DPS

Total DPS = 225.5 + 1476.5 = 1702 DPS

Remember we have 6106 mana left without the MSD, so with the MSD we have to come up with 1655 mana to make it a wash, probably mana spring or two applications of JoW, or drinking a pot/gem early as most of us surely do.

2 Shadow Priests, or a Shadow Priest that does more than 1000 DPS, or JoW, allows this to be sustainable with Bloodlusts/Arcane Power





CRITIQUE AWAY!

I am basing this solely off of the formula on WoWWiki here:
Spirit - WoWWiki, the Warcraft wiki

Please correct me if there is a more accurate one out there.

Mp5 While Casting with Arcane Meditation and Mage Armor = 14.625 + 9 x S / 32

Even assuming the Lock has 10% less Affliction threat (the destro/FG ones would not), you have 100% * -30% (salv) * -10% (talent) * -30% (ranged) = 44% aggro

Vs a Mage with 40% less threat on AE
100% * -30% (salv) * -40% (talent) * -10% (ranged) = 38% aggro


I may be misunderstanding something though, but I know threat redux is multiplied.

Manly, everyone knows AM spam isn't very attractive for raids that don't use JoW. We get that your guild doesn't run a setup that allows you to use AM extensively. Please stop harping on it.

The goal is to find out if AM spam has a place in the raid setting, and the answer seems to be a tentative "yes".

And the above poster was actually very stingy with his maths (though not his assumptions). AM spam w MSD+Ashtongue+TLC+4pcT6 and his admittedly very generous 1800 spell dmg, 33% crit assumptions works out to 1850 dps, pre-Bloodlust/AP at a post-JoW cost of 130 m/second. This is sustainable for 10 minutes assuming a shadow priest, mage armor, BoW, and pots, ASSUMING JoW.

So either my simulator is broken, or AM spam DPS is incredibly difficult to beat in any fight where AB spam is not an option, ASSUMING JoW.

Well sure other specs are probably feasible, even better. I was illustrating AM spam is sustainable.

As for the 433 spirit... I'm at 388 right now with DS, Mark, Draenic Elixir, and Food.

I'm missing Kings, and two items which would give 37 + 10% = 425 + 42 = 467

And I wasn't assuming anything outside the realm of normal possibilities. I was assuming your standard raid buffs, Int/Mark/DivineSpirit/Kings/Wisdom and the Food/Draenic Wisdom consumables, with CoS, Misery, and Wrath of Air. These are all pretty standard in any raid I would think.

If you check my math I wasn't counting Innervates/Mana Tides or JoW to sustain AM.. I merely did the math to show how much mana they all give because I was curious how our mana gets broken down.

I don't think you get my point. JOW or not has no impact on DPS. I just said DPS values in-a-vacuum are more or less interesting. If you give a comparative values, then you can get an idea of whats possible.

And besides, you really seem to view me as some kind of guy that favors one spec over the other. If nobody calls out the numbers, then nothing gets verified, and people just awe at numbers and wild assumptions are made. Heck, it was pretty much 'common knowledge' that bloodlust was reducing the GCD, but it turns out that was wrong.

Sure, in a vacuum it seems that 1700 dps is possible from AM spam, with no bloodlust and no AP counted in. And yes as I admitted even myself, with or without jow you can still maintain it for quite a while, long enough for quite a number of bosses. It doesn't invalidate the numbers in any way. Its quite obvious that any fight thats 2min or less, you will be AB spamming. If the fight is in the 3-4 min range, you have a range of options, which not many have been tested out to be quite honest (besides those that took the time to download vontre's spreadsheet and actually enter a 4min fight duration and played around with it - the spreadsheet tells you directly which rotation is optimal). One of them is AM spam, which I don't believe is optimal under most scenario. Of course fight duration and buffs available will affect the chosen rotation, if you really want to min/max things out. But see, if you assume the same buffs, and the same fight duration, theres nothing in those numbers that indicates that you get better dps or not compared to like, for example, 3x AB, 2x fireball for 70% of the fight and 30% of the time AB spam, which might be very doable given jow for the same duration. Heck, I'm almost 100% positive that you would get better DPS with ashtongue over crusade, even if that would mean worse DPM. Most fights in hyjal/BT don't last 6 min. You can be god damn sure that if I can get 1800-1900 dps from AM spam on almost every encounter (since almost every one of them happen to be below 6min), then I think its pretty darn obvious I'm going to switch on the spot. If mana becomes an issue, then I will probably try AB AM AM. Problem is, I doubt the numbers are very realistic. Oh I sure as hell want them to be realistic. Heck, 1800 dps is starting to compete seriously with AB spam there, which makes it pretty darn obvious that, assuming the numbers hold, theres a great potential, much like I touted myself when the MSD mechanics were different on the PTR.

The real problem is tho, the spreadsheets and my napkin calculations do give me number similar to yours (although quite a bit lower for both AM spam and firespec), but the spreadsheets also tell me firespec should give me almost exactly the same on average, within 20 dps difference. Problem is, I sure as hell wish I could get the dps the spreadsheets tell me I should be getting. AM spam however is a lot harder to 'screw up' than fireball spamming, particularly given its uninterruptible nature, so there are good chances I will get considerably higher DPS because it should be far less affected by latency variance, leading directly to better dps.

And please, don't get me wrong. If frost spec was doing more damage than arcane or fire spec, I would go frost. I don't care what amount of roadblocks you can throw in front of me to make a spec work or not work. I'll take the highest DPS one and that is pretty much it.

EDIT: my bad. I just realize now that when i said 433 spirit was not possible I was thinking out of combat MP5 for some reason. Yes 400+ spirit is very doable.

manly... I think you missed the point of the "illustration" above. It was only meant to show if AM spam was possible over a 6 min fight. as " I wanted to get some more professional feedback on it since people here seem to be knocking Arcane Missiles spam." & "I was illustrating AM spam is sustainable"

he just wanted to know why it is considered not possible to spam AM, so he made the post...
and then he wanted some professional feedback on it.

To me, it looks very visiable to spam AM over a 6 min fight, as ofc this calc, you only rely on a shadow priest. (as the rest of the buffs should be given).

Also, I realy hope that a lot of paladins/the rest of ya, see how much mana jow realy gives. In this simulation you would actualy be able to remove the shadow priest, (the so called "mana battery", the class that has now been whidely used due to its capability to give mana to its group memebers) if jow is up 100%

So Manly, would you please point out where in this calc things aren't right? I respect your posts and I know you know what your talking about (been reading quite a bit of this thread, not all though)


The math I used is as follows:
Arcane Missiles = 740mana + 6% = 784 mana
-56,476 mana (72 volleys in 6 minutes - No MSD procs)
+5,647 mana w/ Arcane Concentration 7.2 volleys @ no mana
+12,000 mana pool
+4,000 mana from pots (2)
+2,300 mana from gems (emerald/ruby)
+7,272 mana from evocate (32.5+0.625x433) / 5 x 2 x 4 x 15 --- 9,090 with 2pcT6
+3,542 mana from Blessing of Wisdom (49*360seconds/5)
+9,821 mana from spirit/mp5 0.45x(32.5 + 0.625x433) = 136.4mp5 x 360 / 5
+18,000 mana from Shadowpriest with 1000 dps (1000*0.05*360)
--------------------------------------------------------------------------------
6106 mana left

+13,320 mana from 100% JoW (72 Volleys x 5 Missiles x 50% chance x 74 mana)

If all those numbers are correct you should be able to switch the shadowpriest for 100% JoW and end up with 1426 mana. ofc I don't mean that you should remove the shadowpriest as there's a lot of "IFs", as when relying on a paladin doing his job to the fullest, I'm just saying "Hey, JoW is almost as good as a shadowpriest, it should be up almost 100% damnit"

I myself can't validate all of the calc above, but most of it looks alright, only thing that I can't say anything about is the mana from spirit/evoc, aswell as the manapool.

I like the upcoming form of MSD because if you proc off the initial cast of AM that doesn't get effected by the proc, you dont have to cancel or cut your slow volley short to get the benefit, focus will be there for you after your slow volley finishes.

Just wanted to point that out.

Also to say, my numbers are inflated because as I noticed last night, AM spam incurs an inherent delay between volleys. After the last Missile of volley 1, and casting volley 2, there is like a 1s delay between Missile damage. So its not really 360 Missiles in 6 minutes.. Its gonna be closer to 280 which severely effects the Theory Craft.

Or 360 Missiles in 7 minutes, 12 seconds.

I've tried to stopcasting the volleys but it still robs me of my last Missile.

12,000 mana raid buffed is very feasible in the end game gear we are talking about. Even more for those damn gnomes.

The evocation math I got like this:
+7,272 mana from evocate (32.5+0.625x433) / 5 x 2 x 4 x 15 --- 9,090 with 2pcT6

Evocation allows 1500% regen while casting for 8 seconds, ticking every 2 seconds, so 4 ticks total.

32.5+0.625x433 = 303 Mp5 idle / 5 = 60.6 Mana per Second

60.6 x 2 = 121.2 Mana per 2 seconds

121.2 x 4 ticks = 484.8 Mana per Evocation w/o 1500% bonus

484.8 x 1500% = 7272 Mana

2pc T6 gives you an extra tick, so multiply by 5 instead of 4 = 9090 Mana

That's because you assume the wrong things. AM does not work like most people believe, ie:

AM DOES NOT WORK THE WAY DESCRIBED BELOW
---------------------------------------------------------
Every 1s (which is affected by haste/slow effects) starting from when AM started channelling will send a bolt. This means that 'refreshing' the AM channelling every 2.5 seconds (illustrative number) should lead to 100% DPS time. At every precise second a bolt will shoot.

AM WORKS THIS WAY
------------------------
Every 20% of AM, a bolt will be sent. This means that 'refreshing' the AM channelling every 2.5 seconds (illustrative number) will shoot 2 bolt per AM, and the extra 0.5s will be lost.

I did try to mention it in my previous post, unfortunately there is no way to bypass lag even with excess mana spent at the problem, even on channelled spells.

Completely agree about the way MSD will work in 2.2, I really think it is amazing and a lot better then the current having to cancel your current channel to utilize the proc, even if you can get 2 AMs during that proc.

As for the /stopcasting the channels and all of that, just remember that there is the same lag inherent with all spells even if you use /stopcasting. Best I was ever able to average using /stopcasting was 2.67 seconds average frostbolt before I started clipping them occassionally, and that was just on a couple dozen mana bars spamming spells at Dr Boom under real raid conditions I think you would be lucky to get that close to actual casting times. Arcane Missiles actually suffers less then every other spell since it is a 5 second channel and longer cast/channel times means that latency has less of an overall effect.

Besides hopefully that little lag will become a moot point in 2.3 with the latency fix (whatever that happens to be).

Well it effects Missiles worse I think because it seems to take a second before you begin casting your first Missile.

So instead of modeling 5 missiles per 5 seconds, its really more like 5 missiles per 6 seconds... making it an effective 83% efficient.

I can chaincast FB/FrB/AB without this problem, as its part of their built in cast times.