What were you using to establish the 1 crit = 1 damage metric? I question this for two reasons. Firstly, and most obviously, crit scales the 2 arcane trinkets of choice: Ashtongue Talisman and Lightning Capacitor. I know you considered the lightning capacitor, but you did consider Ashtongue Talisman uptime? I Know for a fact my uptime is definitely not 100% as I'd like it to be.

Secondly, crit scales directly with +damage. For instance, at 800 spell damage, I might do the math and come to the conclusion that 1 crit equals .5 spell damage. But if I redo the math at 1200, I would have very different results. The more +damage you have, the more bonus you gain from adding more crit.

How did you establish what +damage to consider as the "base"? I know I personally sit at 1620 arcane damage with full raid buffs/consumables/wrath of air. At that level of +damage, and both my trinkets, I think I've reached the point where I can value crit a bit higher than I could previously.

He's saying in cases where you might not be able to count on 100% JoW uptime or be in any other less than ideal conditions that the value of spirit goes up. Making Devastation better for some.

I'm usually a fan of spirit and I love my spirit on my gear, however one thing is theory, another thing is reality, it's rare that fights go on more than 8 minutes, out of all the tier6 content 2 fights are longer than 8 ish minutes by their design? Council and Illidan. Also it's worth keeping in mind that as your guild improves/progresses fights will be shorter and shorter and the relative gain from spirit deminishes. I still wear my devastation bracers and t6 shoulders over the haste ones though, just because I like getting 100% mana from 0% with evocation - stupid 2.3 changes will be a nerf to my lovely nightbane staff.

Of course it's not fixed, but I don't think I used a situation where spirit is worst at all.
With molten armor for example, Spirit is much worse than with Mage Armor. Also, the average fight isn't 10 minutes of spamming Arcane Missiles, so spirit is less useful in most of the fights as well.

I think the situation I used is a pretty good estimate of what will be best on average.

Also, he said he values spirit for 10 minute fights as around 2 spell damage per point of spirit. I DID use a 10 minute fight so Spirit shouldn't be undervalued in the calculations, should it?

Edit: @Faxmonkey:
I got the value for 1 crit rating = 1 spell damage from one of the posts above but I can try it out in the spreadsheet.
Edit2: Switched some gems around and came to the conclusion that 1 crit rating = ~0.8-0.9 damage for the gear listed above.
Edit3: About spirit: In general I'd say as long as you don't go totally oom in a fight you shouldn't bother with Spirit at all as it's of no use for you.

A totally different thing I'd like to ask, does anyone have any experience with other speccs than Deep Frost in high rated arena (for example Arcane)?

I tend to agree about spirit being something I rarely choose over anything else. Yes, the conditions you stated would make spirirt pretty undesirable, but I think he meant something more like little / no JoW uptime and a mediocre sPriest, or none at all, in a situation like this someone that already had mana trouble might give slightly more consideration to something with spirit vs. something without.

Changed the values around a bit, with a 100 MP5 Priest (that's only 400 DPS, my level 64 twink does more than that) and only 50% judgement uptime the Cuffs of Devastation are equal to the Bracers of Nimble Thought with Molten Armor up and worse with Mage Armor up. In general I'd still say, the Bracers of Nimble Thought are better in nearly every situation with the Cuffs being a good alternative.

I can hopefully do this on Monday. I'll probably use a sim rather than math, since MSD, arcane potency and ashtongue itself add a lot of variables to the equation.

I've been using Vontre's Mage DPS Spreadsheet mostly for final dps calculation which includes a calculation for Ashtongue uptime so the numbers should be about right. As it includes buffs and buff food/consumables as well there shouldn't be any big errors in there.
However a pure Arcane Missiles DPS calculator shouldn't be too hard to write anyway as it's pretty simple. (Nothing like Arcane Blast rotations in there)

If you're really using a sim you should be aware that you need to run a very large simulation to get useful results as you already stated there are many random variables that can totally change the output.

Editing.

Yeah, it can take a painfully long time to do. Hopefully, the results will be able to be plotted on a line allowing us to cut through some of the noise.

I understand why you are using your arguments but I disagree that spirit is as poor as you make it out to be for arcane mages.

Your claim is that 21 spirit (23 spirit for human), Kings included, is worse than 13,8 spell damage. This is wrong (On certain boss encounters yes but on average, no). 21 (23) spirit is better than 13,8 spell damage for an overwhelming amount of arcane mages (SSC/TK/BT/HY) lvl.

That makes the Cuffs better than the Nimble any way you want to look at it as a primary item (And makes every other item with spirit on it considerably better than ppl think). If you have both, then ofc, switch to it on certain encounters, but how many ppl do you honestly think have both the cuffs and nimble.

You cant look at spirit in a "Teron Gorefiend" perspective (Extreme) when there are encounters like Kael`thas, Archimonde, Illidan, unforseen problems, no JoW uptime, deaths (Dead SP anyone) even trash packs. You have to "average" out situations when calculating spirit and you have to include a certain number of full MP5 ticks.

I have to agree with Blizz`s calculations regarding spirit. Its the most underrated stat for us because most ppl don't grasp the complexity of it (Raiding is more complex than a Dr Boom/Teron encounter).

And with 2.3 coming, the value of 1 spirit will increase substantially for arcane mages.

I do agree however that if your guild has Illidan on farm, 50-100% JoW uptime, an SP that never ever dies and is in your range 100% of the time, then the value of spirit will significantly decrease. But seriously, is it realistic to calculate gear value based on these factors?

Keep in mind that less than 1% of all SSC/TK/BT/HY guilds is in a situation like this.

It depends very much on the length of the boss fight and the Shadow Priest as well, with a good Shadow Priest the Bracers of Nimble Thought are better on most of the fights, without a Shadow Priest and JoW, the Cuffs of Devastation clearly win on most fights, however under these conditions Fire wins over Arcane as well - after the patch even more than now.

Under some conditions it's also better to use AB, 2x AM. This cycle lets Spirit really shine as you get a normal spirit tick on every AB.

In conclusion, get both and put on the Bracers of Nimble Thought if you don't have any mana problems on the fight and the Cuffs of Devastation if you do.

That complexity can, however, also hurt Spirit. For example if there's a 15 minutes fight where you stand around 5 minutes doing practically nothing. If you were able to spam Arcane Missiles for exactly 8 minutes until going oom under the given circumstances, the 5 minutes of doing nothing (and letting your normal spirit regeneration tick) allow you to easily spam Arcane Missiles every time you are able and allowed to. The Bracers of Nimble Thought would be better here if your normal Spirit is enough that you can spam Arcane Missiles every time you should. The Cuffs of Devastation would probably be better if your normal Spirit regeneration was that bad that you could only spam Arcane Missiles for like 6 minutes on a Gorefiend like fight and maybe 8 minutes of this fight, letting 2 minutes of possible DPS being wasted.
Look at my conclusion above, I think it's right as of now.

I was wondering if it would be possible for an estimated JoW uptime on a typical BT/Hyjal run, for bosses at least, for you guys guilds.

The concensus in mine at the moment is "JoW's too difficult to maintain, drains too much mana and isnt worth the effort" and as such see basically 0% JoW uptime.

I assume they'r just over-reacting right? Ive noticed now moving deeper into BT that without JoW AM spam is getting rather difficult to sustain and would appreciate anything which can help convince them they'r wrong in this matter.

That's the consensus aswell in my guild and I can understand them actually. Usually we only have two arcane mages in the raid and jow really is only awesome for us and to a lesser extent, hunters.

In contrary to what others are saying here, I find it hard aswell to keep the mana up without jow. I always get a spriest of course but I have to use mage armor and chain mana pot most of the times, I'm not even close to the regen where I could use molten armor.
Jow would allow me to use molten and maybe destro pots but it is really worth the effort to give two mages 100-200 dps more?

On the other hand, maybe you have quite a few ferals druids that never use their innervate or if you have some really friendly druids, they innervate you without even asking for it

The only fights that our Paladins have conceded that they can keep up JoW on in BT/Hyjal are Teron Gorefiend (everyone is clumped up on top of him or near him) and stage 1 Essence of Souls (they can't heal). Every other fight there's a reason why we can't get JoW. I'm curious what other guilds in BT/Hyjal experiences with JoW are?

Fortunately I haven't found JoW to be needed at all with AM spam with a Shadow Priest and Tide Shaman. Although I do have a tendency to use Mage over Molten Armor. If JoW had 100% up time I think it would be possible to go without a Shadow Priest which could potentially help the raid quite a bit, either by giving the Priest to healers or replacing the Priest with a class that outputs higher damage.

JoW is something of a rarity for us at this point but I keep hammering on our pallies about it. The irony is that we have great uptime on farm-status fights (familiarity breeds bored paladins apparently) but when we could really use it, we often just do not have it. It certainly is possible to put it up on many, many targets though and it really shines through when used 'properly'. I try not to ride them too hard about it though as obviously they are busy people with other buttons to mash and those buttons tend to be keeping our tank(s) alive. I briefly brought up the theoretical idea of one paladin (we raid with 3 or 4 usually) speccing ret in 2.3 and was shut down in a hurry. Our pallies are all rerolls naturally and they rolled the class to heal quite clearly.

Running arcane and still with the T5 bonus, more mana is always a good thing for me. I can definitely sustain AM spam (no Ashtongue) off just a spriest with pots and could probably sustain it off just JoW. I'll take every drop of mana I can get though as at present AB spam is still more damage than AM spam.

Someone have theorycrafted about arcane vs fire post 2.3 cofficent tax removed?
Arcane will still be better than fire?

Our pallies try to get JoW as much as possible from what they tell me, and just like you Zwinkz, it isn't much more than the fights that you listed.

'JoW is a mana drain' is said by terrible paladins. Feel free to tell them that. Maybe they don't know that melee attacks refresh it. Even if they're holy.

Our JoW uptime is fairly high even though we don't currently have a ret paladin. A few fights require a little bit of skill to keep it up. But if you're playing at this level, you better have a little skill.

It should be noted that pairing Counterspell with arcane missles in a macro can increase dps because the counterspell can proc focus while not consuming it

Has anyone tested out how good [Darkmoon Card: Blue Dragon] would be with AM spam.

This could be insane mana regeneration, coupled with 2.3 Arc Meditation, mana would never be an issue.

Are you sure about this? Because the tooltip clearly states "after any damage spell".