I believe Arcane Missiles has an "invisible" debuff slot. You can see this on Houndmasters in Heroic Shattered Halls, who periodically remove all debuffs on themselves. People have reported that it will stop an AM in midcast. The AM debuff was also noticed back in the 16 debuff era, but that was pre-2.0.

I think all channeling spells work through a debuff. That's why they'll keep damaging their target after a mind control causes them to be friendly to your side.

Your posts starts quoting what you consider a truism, but which has in fact been proven a few times (last I recall Roywyn posted in [Mage] TC after 2.3) to be wrong.

Fire Blast starts significantly ahead of Fireball, but it scales very badly in comparison and the comparative advantage combined with 4% crit difference only carries so far. After a certain gear threshold which isn't too hard to get to (think T5ish) it reaches a point where Fire Blasting starts being a DPS loss. Coupled with it's atrocious mana cost, what's widely held is that Fire Blast is only good for moving and for Combustion Double-Dipping (even though it's not 100% guaranteed that that's a good idea to boot, as on occasion the Fire blast ignite might overwrite the Fireball's one netting in a 1.5sec wasted with a DPS loss).

As for your question of haste: Yes, I think it's rather evident that when you spend 1.5sec out of your 7.5sec rotation doing a spell that can't be hasted, haste loses a good chunk of it's value.

This also holds true for the gentleman who posted before, regarding haste "losing" value the more he stacked with arcane-fire spec. It's not "losing" value, the simulator is merely giving you data you're interpreting wrong. Given 0 haste, GCD is not preventing anything, hence haste "should" increase total DPS by, say X. When you in fact gain 1 haste, half your spells gain GCD prevention from increase, resulting in Y (>X) increase. This only means that the simulator was displaying X, given the circumstances would not chance ie. given you'd not run into GCD. Seeing as you do, the prediction of X is wrong.

The fact that you're seeing multiple numbers of haste relative benefits just means there are "staggered" values of haste which reduce themselves each time a spell in your rotation (ie. first is the 1.5sec AB, then the AB before that etc) hits the GCD barrier.

The fact remains to both posters: You're both advocating/referring to stacking a lot of haste; this suggestion has been floated in the past but falls short on the fact that almost all the ZA/badges haste gear has a substantial lacking in anything except stats and +dmg. That being the fact, while stacking hypothetical haste might work, it will not provide substantial benefit over losing a lot of hit rating. It's not so much a case of "haste doesn't work" as a case of "most specs can't deal with haste gear" (or more accurately "deal with the loss of non-haste gear")

One thing I'd like to add: What's the only spec that doesn't particularly care about crit and hit? What's the one spec that will minimise the loss of "relative" benefit from haste because it hits once per second? That doesn't need to wait that "25th fireball" to get the benefit from 4% haste? Arcane.

Haste gear is perfect for arcane. The item stats are nice and well-placed (plenty of int, plenty of +dmg) certain parts have spirit on them (ZA chest hast plenty), there isn't overflowing hit rate, and there's still space to get as much as you can and have 2 parts free for T5.

Note however, I'm not advocating arcane spec, on the contrary. I'm providing an illustration as to how haste gear would be in it's element so to speak (droll pun intended).

Meh, I was wrong on this one. A Warlock guildmate pointed it out to me.

Using the formula of (Non-crit damage * non-crit percent + 1.03 * Crit percent * crit damage) / (Non-crit damage * non-crit percent + Crit percent * crit damage) makes it a logarithmic curve. Which means that the percent increase CSD gives suffers from diminishing returns (but is still growing) as a person's crit percent increases. The simple math I used matches at 0 and 3. As you can see from my chart, slightly less than 20% crit rate makes the CSD worth about a 1% increase in damage, 50% crit rate worth about 2%, and 100% crit worth 3% increased damage.

Here's the numbers for a deep fire (210% damage on crits) percent increase from CSD. Deep Frost (200% crits) is close enough for eyeballing, getting from 0.01% to 0.04% less of an advantage until they converge at a 100% crit rate.


I still prefer the Cowl/Hood of Hexing over Tier 5 helm- but that's mostly because Tier 5 has the useless Spirit on it.

Only, for frost, the CSD pushes the crits from 200% to 209%, a 4,5% gain, due to funky mechanics (the formula: CSD crit multiplier = 1 + (1,03 * (base crit multiplier) -1) * (1 + talent modifier) = 1 + ( 1,03 * 1,5 -1) * 2 = 1 + 0,545 * 2 = 2,09.

For Arc/Fire, we go from 245% to 254,45% crits, a 3,86% gain.
For Arc/Frost, we go from 225% to 236,25%, a 5% gain.
For Arcane, we go from 175% to 181,75%, a 3,86% gain.

(note: Ignite is calculated after crit damage, which is why it doesn't double-dip on the CSD like the other crit talents)

How sure are you about this? I don't really raid frost much (save immunities), but I remember a long time ago reading a huge analysis showing that Frostbolt and binary spells alike recieve an added resist chance proportional to the %damage reduction to non-binary spells from level-based resistances. I would think that amidst all the recent testing done on frostbolt's hit% with respect to elemental precision that the answer should be pretty obvious whether what you're saying is true or not (which suggests that you're correct). However, it is also quite obvious that resistance levels in PvP and level-based resistance rates in PvP do affect binary hit chances which would suggest that level-based resistances should cause additional resists to binary spells. I've just always thought the latter is how it works and was surprised to see you say binary spells are unaffected.

Also, as an addition to the list of bugs you may want to include the AOE+MoE bug.

Yeah, you're right.

Here's a nifty chart I made!



Values I used-

Deep Frost - 209%/200%
Deep Fire - 216.3%/210%
Deep Arcane - 181.75%/175%
Arcane/Fire - 254.45%/245%
Arcane/Frost - 236.25%/225%

The percentage increase in DPS is due to Chaotic Skyfire Diamond is easy to generalize: namely, for crit chance c and crit bonus b (note that b = 0.5 without talents, not 1.5), the percentage increase in DPS is given by...

Note that ∆b = .09*b without Ignite or .09*b-.036 (which is just .09*(b-.4) with Ignite. These expressions yield the correct values for the new crit damage--for example, ∆b = .09 when b = 1 (Ice Shards), meaning your new crit bonus is 109%, as has been determined.

%DPS increase without Ignite:
With Ignite:
I'll spend some more time later trying to come up with a more succinct form of the latter expression.

As for whether binaries can be resisted partially: Binary spell - WoWWiki - Your guide to the World of Warcraft

If you prefer circumstantial evidence, I've raided frost almost exclusively, appart from the first karazahns we did (33.28.0, woe is me) and 2.2 arcane AM spam. Not once ever have I noticed a number that wasn't within 120dmg of my average (and thus within the range on tooltip). Debuffs on mob, trinket procs and trinket activations, destruction pots and Spellstrike Proc where appropriate, it goes up. Less than 100% unbuffed non-crit value, I have never seen. Partial resists, unless I'm badly mistaken, are 25, 50 or 75% and I've never analyzed data that gave a minimum bolt hit of less than circa 1400-1600 (depending on gearing at the time).

Keep in mind, it's a whole lot easier to keep track of the number-spam in frost: There's no ignite DOT going off, no fireball DOT going off, and no scorch every-so-often to throw you off the number count; A wrong number in frost stands out like a sore thumb extremely intensely. The focused "trance" of awareness you fall into when you're frost has a lot more attention tuned to the numbers that come up because they're less prone to be all over the place like fire is.

Is the General Consensus that 2/48/11 is going to be the best pure dps spec in 2.3.2. If not how much will Icy Veins affect deep frost (3/0/58), compared to this new Fire/Icy Veins Spec(2/48/11). I was wondering if Vontre or one of the other theory crafting mages could help me out. Thanks

It was actaully a comparison of deep fire vs deep arcane not arcane-fire. I could understand why haste would be valued so low for Arcane since I could easily see how Ab rotations would devalue haste vs +dmg (although I was surprised to see it drop in relative value below crit); however, according to the numbers the script gave me, the relative value of haste was below +dmg for deep fire too once you had more than 1 piece of haste gear (lets guess 20 points because I forgot and eitherway several times that amount of spellhaste is relatively easy to get). Since deep fires rotation is largely immune to GCD clipping THAT is what made me wonder what was going on. Unless that one scorch in a 30s cycle can really mess up the value that much I figured that these numbers weren't matching what I kept hearing. I did hear someone mention that there is a +dmg threshold (~1100?) where haste gains superiority to +dmg. Frankly.. i'm going to have to model the equation myself to figure out exactly how it works and I thnk I've learned enough to do so. However, the point still remains taht spell haste does seem to shift in relative value much more quickly than other stats I've seen... but there could be several reasons for that... including an error in method I may have made while running the script. Insofar as that its true tho, when min/maxing armor it just becomes a pain to work in spell haste gear for me which is what i've been trying to figure out how to do for the past 2 weeks.


Refering to my original post... the numbers for spellhaste showed it below even crit in value. For AM spam... it would be excellent... but having to work in an AB rotation would make it almost useless.

New EJ poster here; a bit intimidated by the number-crunching and "elitist" (pun aside) mood around, but I do have a question pertaining on the future of proper mage raiding spec, preferably past Karazhan.

Right now, I am your classic ABx2, AM, Scorch mage with 48/13 spec, and I keep hearing that the rotation is not recommended anymore due to 2.3's various nerfs and buffs to other trees than Arcane. While I love the playing style of Arcane, I'd like to switch to a more favorable spec in terms of maximum DPS output.

Time to time, I keep hearing about the new acceptable raid rotation as an Arcane/Frost spec: AB x2, Frostbolt x2, repeat. I assume it's because of the 225% damage on Frostbolt crits due to Spell Power and Ice Shards combo, and due to the 10% tax repeal. Is this information true in its most basic form, and can it exceed my current spec in terms of damage output?

PS: In terms of gear, I have no problem in reaching (back) 164 Spell Hit or whatchajigger less due to Elemental Precision bug. Also, if WoW Armory hasn't updated yet, I picked up [Blade of Twisted Visions] from our good ZA run last night.

PPS: Yes, that's my raiding gear.