If all 3 mage trees were equal, hardly anyone would be going fire - frost has survivability, and arcane has more toys. I agree with others that the 3 mage trees are largely unique but have their purposes - arcane can burst like nobody's business, frost is survival, and fire is pure consistent dps.

This thread isn't supposed to be Arcane vs. Fire or whatever it's turned into now.


To get back to the T5 bonuses, I am assuming the 4 piece bonus has no internal cooldown, can anyone confirm that it can refresh itself?

On Arcane Blast

I was tempted to write about some of my own experiences with 40Arcane/21Frost (specced for Al'ar first kill) and getting 2/5 T5 afterwards, but decided to skip it not to incite more emotional outburst.

There still are a couple of things I'd like to talk about/suggest thinking about:
- Using Arcane Blast doesn't equal OOMblast spam. Yes, it it situationally useful to blow a full mana bar on 10% of a boss fight (I did that on Leotheras in the last 15% phase, used Evocation on the transformation and went OOM right before he died).
What I am personally interested in is sustained DPS, where casting Arcane Blast with 3 debuffs on (when the cast ends) is a big no no. I felt that I had decent DPS with normal mana consumption from that week's experience.
- There are many occasions where I have to stop nuking at some points. Silences (Gruul), running out of AoE (Gruul, Al'ar, Void Reaver, Leotheras), repositioning and running after adds (erm ... every fight?), watching and repolying sheep (trash, Lurker).
You can time quite a few of those, so that your break your casting while the debuff cools down, and you only loose your filler cast, not your Arcane Blast casts.
No, it doesn't always work.
Yes, fire is at an advantage there with 6/12 yards higher range.
The point is that the fights consist of more than just cycles that you can easily compute.
- On our Vashj kill, mages were mostly strider killers, helping on naga and elementals, and standing guard at the top of an assigned quadrant when the timer for tainteds is up.
I used Blast on elementals, blast/frostbolts on striders, and 1 or 2 fireballs on emergency elementals. I spent a lot of time running for/after things, which could be timed really well for the debuff to fall off.
This is way too unpredictable to model.

I'm hoping to get the talk away from people fixing their eyes on OOMblast spam somehow :o


DPM question

Back to something more substantial to talk about.
DPM with various specs and OOM time.

From my personal spreadsheet, Fireball rank 13 (level 66) has the highest effective damage per mana from all specs/spells, the unimproved one actually beating the improved one by 0.5%.

I don't know how other spreadsheets work, but in my opinion, the amount of mana not regenerated while casting should always be added to the mana cost for the spell. I actually overdid it by reducing the spell's DPS by my wand DPS, because that's what DPM boils down to for me:
If I'm completely OOM, I wand and regen some mana back (I managed to avoid that so far mostly).
Casting a spell would give me a damage increase (difference between spell damage and wand damage) at a mana cost (spell cost plus the mana from spirit that I would have regenerated if I weren't casting).

From this point of view, nothing beats Fireballs (10/48/3) in my sheets, so, I'm at a loss how everyone else claims arcane has superior DPM/longvity.

If scorch spam lasts 8 minutes, fireball spam lasts 5 minutes, then the highest damage I could get out of it would be fireballing, wanding for about 2 minutes and then fireballing again. Scorch feels more mana-efficient, but fireball is actually more efficient due to its higher DPM.

Please tell me I'm totally wrong about effiency or so :o


Prediction/analysis on spell haste

If we look back, spell hit is a rather new stat, introduced with BWL/ZG. Its triumph was that it offered a lot of benefit for low item budget, at the expense that it would only affect bosses. Spell hit was better returns per budget than +damage, and way better than spell crit.
So, many items with hit and damage (or hit and crit and damage) were way better than those without hit, which is way we have a lot of spell hit floating around now.

Looking at spell haste, it's a bit better than crit (1 crit rating ~ 0.67 dmg, 1 hit rating ~ 1.5 dmg, 1 haste rating ~ 0.8 dmg for 10/48/3 with +1150 dmg) for pure DPS, and makes you drain mana faster.
So it's marginally better than crit if mana is no concern, and way worse than hit unless you're hit capped.

For DoT classes that have next to no use for crit, spell haste would have provided a second useful stat to allow making better gear by mixing several stats. Sadly, it doesn't affect the GDC and is pretty pointless for exactly those classes.

To be competitive/superior for casters that have a use for spell crit/hit, items with haste would need to have hit and crit as well, to get more points out of item budgets. The current spell haste gear doesn't, and is mostly a sidegrade/slight downgrade to gear of the same tiers.

At least, that's what I got when plugging in the spell haste gear so far, most are DPS sidegrades, and downgrades if you're not hit capped.
Some are craftable though, so they make for decent filler gear when some item slots just refuse to drop.

So, from theorycraft, spell haste is okay, but not well enough itemised to provide DPS upgrades, especially when mana is limited.

A few percent of spell haste can actually be useful for arcane if people have problems casting their filler spells between arcane blast cycles before the debuff falls off.

No CD, can refresh. That's how it worked on the PTRs at least, and no item set bonus proc so far has an internal cooldown either.